On Wed, 30 Jul 2014 19:10:04 +0200 Nicolas George <geo...@nsup.org> wrote:
> Le duodi 12 thermidor, an CCXXII, wm4 a écrit : > > You never explained what's wrong with it. > > This is tiring. Everyone knows it already and it has been explained numerous > times. > > THERE IS NO TIMESTAMP FOR THE END OF STREAM. You could just have said that right in the beginning. Even in the commit message you could have made explicit that the purpose of the change is adding an end PTS, instead of that being a positive side-effect. > Is it enough, or do I have to waste more time explaining how that makes the > duration of the last frame wrong, and how that is a problem when accurate > timing is necessary, for concatenating or looping for example? > Demuxers and decoding don't return such an end PTS, so why should filters? Demuxers apparently use the frame duration to time the last frame, so why not use that in lavfi too? (Demuxers also report the total length, but that is probably not useful at all for timing the last frame.) _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
